As it stands, the question and answer are completely specific to this one song, so it seems like it's not useful to other users.
I'd agree that It isn't useful to other users in the sense that other users are very unlikely to come to the site with the same issue with the same song. It's also perhaps not all that likely to be a highly-ranked search hit.
However, it could be useful in the sense that users who do come across the question might find it an interesting case study, and someone (or maybe more than one) might gain satisfaction from answering. In this case the number of upvotes signify to me that this particular question is useful to other users, in this sense.
Would it be better if the question were more along the lines of, "when trying to re-create a song, how do I figure out what I'm missing?" so it's more general?
I think the lack of specifics would make that too broad. Often the devil is in the detail.
Is this on-topic?
I think I can see why you ask. This kind of question is somewhat close to something like "here's my original composition/my performance - how did I do?", which I don't think is usually seen as on topic, and also touches on the "Identify this sound" type question, which we decided to kick over to Sound Design.SE.
As a site we seem to have decided that we're wary of questions that don't become easily-searchable additions to the "knowledge base", so we slightly lean away from the "this is my very specific problem" kind of questions that some SE sites do try to tackle.
From what I can see, the user base here often seems happy to leave things fuzzy and take things case-by-case. If that's true, then I guess the upvotes mean that this one is fine..?